International Journal of Food Microbiology 139 (2010) 15–22

Contents lists available at ScienceDirect

International Journal of Food Microbiology

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j f o o d m i c r o

PCR multiplex for detection of Salmonella Enteritidis, Typhi and Typhimurium and
occurrence in poultry meat
Camila Guimarães de Freitas a, Ângela Patrícia Santana b,⁎, Patrícia Helena Caldeira da Silva c,
Vítor Salvador Picão Gonçalves b, Márcia de Aguiar Ferreira Barros b, Fernando Araripe Gonçalves Torres b,
Luci Sayori Murata b, Simone Perecmanis b
a
The Animal Health Programme, Faculty of Veterinary Medicine – University of Brasília, Brazil
b
Faculty of Veterinary Medicine – University of Brasília, Brazil
c
The Animal Science Programme, Faculty of Veterinary Medicine – University of Brasília, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: The occurrence of foodborne diseases is increasing throughout the world. Bacteria of the genus Salmonella
Received 11 September 2009 are responsible for food poisoning and, in some cases, may be fatal. The aim of this study was to adapt the
Received in revised form 14 January 2010 multiplex PCR technique (mPCR) on the rapid and direct identification of the presence of Salmonella sp. as
Accepted 7 February 2010
well as serotypes Enteritidis, Typhi and Typhimurium in poultry carcasses (n = 127) and viscera (n = 73).
The implementation of the standard technique using positive controls was successfully adapted. The results
Keywords:
of Salmonella sp. detection in refrigerated viscera showed that the mPCR was able to detect Salmonella genus
Salmonella
Enteritidis
in 2.74% of these samples. Traditional microbiological analysis also identified the same positive samples for
Typhi Salmonella sp. but was not able to differentiate the serotype. The serotype Enteritidis was detected by mPCR
Typhimurium in 1.37% of the samples. Our conclusion was that the mPCR was able to detect the presence of these bacteria
Poultry in a short period of time and enabled the identification of serotype Enteritidis in one of the samples found
Multiplex PCR positive for Salmonella sp.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction et al., 2007). The technological procedures for obtaining chicken

Salmonellosis is one of the major foodborne diseases. Due to its
endemic nature, high morbidity and association with a wide range of
foods, this zoonotic disease is of high public health concern (Aarestrup
et al., 2007; Alizadeh et al., 2007; Kottwitz et al., 2008). Outbreaks of
salmonellosis are rarely reported in Brazil (Kaku et al., 2000). The
prevalence of these bacteria in different types of food is thus
underestimated (Kottwitz et al., 2008). Salmonella serotype Typhi is
human specific (Arya and Agarwal, 2008; Ochiai et al., 2008) and
when found in food indicates handling and hygiene failures (Chen et
al., 1997; Che et al., 2000). Furthermore, the serotypes Enteritidis and
Typhimurium can be isolated both in chickens before slaughter and in
humans infected by foodborne diseases (Rabsch et al., 2001). Non-
typhoidal salmonellosis is a major cause of human gastroenteritis
worldwide (Kornschober et al., 2009). Kaku et al. (2000) reported 211
infected people during an outbreak of Salmonella Enteritidis in a São
Paulo city school caused by food containing raw eggs.
It is considered that the presence of Salmonella sp. in chickens
makes it unsafe for human consumption (Bjerrum et al., 2005; Agunos
⁎ Corresponding author. R. 15 FAV-UNB, Postal code: 4508, Asa Norte, Brasília – DF,
Brazil. Tel.: + 55 61 33072823; fax: +55 61 3273 6593.
E-mail address: patvet@unb.br (Â.P. Santana).
carcasses and cuts for consumption are also potential risks for
contamination (Deseo and Engeli, 1979; Christensen et al., 1997),
especially in evisceration, cooling, packaging and transport stages
(Rasschaert et al., 2008; Muth et al., 2009) where microbial growth
can occur.
According to the Brazilian Association of Chicken Producers and
Exporters (ABEF, 2008), Brazil is the third largest producer of chicken
meat, (behind the United States and China) and is the main exporter,
followed by the United States and the European Union.
Salmonella sp. microbiological research in poultry meat involves
stages of pre-enrichment, enrichment and selective plating as well as
biochemical and serological tests (Foley et al., 2007). Negative results
are obtained in about 4 days whilst positive confirmations normally
take up to 15 days (Blixt et al., 2007; Gallegos-Robles et al., 2009). The
conventional method for microbiological diagnosis in the case of this
pathogen is also checked by the fact that isolation requires a sufficient
number of viable cells (Chaicumpa et al., 1995). Moreover, morpho-
logical changes and the biochemical profile of the colonies both
increase the uncertainty of this diagnosis (Fach et al., 1999). The
multiplex PCR (mPCR) is a variation of reaction chain polymerase
(Kumar et al., 2006; Kumar et al., 2008), which involves more than
one pair of primers allowing one single reaction to detect more than
one type of microorganism along with their serotypes and/or different

0168-1605/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijfoodmicro.2010.02.007
16 C.G. de Freitas et al. / International Journal of Food Microbiology 139 (2010) 15–22

Table 1
genes, therefore reducing the diagnosis period (Malkawi and
Salmonella sp. serotypes, the target gene, sequence and the size of fragments to be
Gharaibeh, 2003). amplified adapted from Kumar et al. (2006) and Alvarez et al. (2004).
Considering the importance of this group of bacteria in the
processing of chicken, its importance to public health and considering Bacteria Target Primers Sequence Size Access
gene 5′–3′ (bp) number
the need of fast and accurate diagnosis methods, we worked towards (GenBank)
adapting a multiplex PCR to detect genus Salmonella and serotypes
Genus Salmonella ompC OMPCF atc gct gac 204 AY341077
Enteritidis, Typhi and Typhimurium in refrigerated carcasses and
spp. tta tgc aat cg
chicken viscera. In addition, we were able to estimate the occurrence OMPCR cgg gtt gcg
and frequency of these microorganisms in chicken in Federal District, tta tag gtc tg
Brazil. Salmonella Enteritidis Sdf I ENTF tgt gtt tta tct 304 AF370707
enterica gat gca aga
serotypes gg
2. Materials and methods ENTR tga act acg
ttc gtt cttctg
2.1. Positive controls g
Typhi ViaB ViaBF cac gca cca 738 D14156
tca ttt cac cg
Positive controls were obtained from the Oswaldo Cruz Founda-
ViaBR aac agg ctg
tion (FIOCRUZ), located in Rio de Janeiro, Brazil, which is a reference tag cga ttt
center for Salmonella serotyping. The strains from FIOCRUZ (Salmo- agg
nella enterica serotypes Agona, Enteritidis, Typhi and Typhimurium) Typhimurium Spy TyphF ttg ttc act ttt 401 AE008757
were plated in blood agar and then McConkey agar. Subsequently the tac ccc tga a
TyphR ccc tga cag
bacteria were inoculated in 5 mL of liquid L broth containing 0.5% (w/ ccg tta gat att
v) yeast extract, 1% (w/v) casein peptone and 1% (w/v) sodium
chloride. Each medium was then transferred to a 50 mL Falcon tube
and homogenized overnight at 37 °C in a shaker (New Brunwick
Scientific, Edison, N.J., USA) at 200 rpm, from where DNA was
extracted. (1.5 mM, 2.0 mM, 2.5 mM, 3.0 mM and 3.5 mM). The average
annealing temperature of primers was estimated at 57 °C, but for
2.2. Extraction of total DNA from the positive controls Salmonella Typhimurium the annealing temperature was 55 °C.
The PCR was performed with 10 pmol of forward primer and
Total DNA extraction, in this case, Salmonella's strains that were 10 pmol of reverse primer, corresponding to Salmonella genus and
previously inoculated in liquid L broth, was performed with phenol, each one of the serotypes proposed. The mix of reagents was
chloroform and isoamyl alcohol (15:14:1), precipitated with 100% composed of 3.0 mM of MgCl2 (Cembiot®-RS), final 2.0 mM of
ethanol and sodium chloride (0.3 mM final concentration), following dinuclueotide (dNTP Amershand®), buffer 10× (Cembiot®-RS) in
the recommendations of Sambrook and Russell (2006). The result the final concentration of 1×, 2 U of Taq polymerase (Cembiot®-RS),
from each serotype was quantified in 0.8% (w/v) agarose gel with 5 ng of DNA as a template and MilliQ sterile water for completing
0.02% (w/v) ethidium bromide using high-weight molecular mass 25 µL of reaction.
ladder (Invitrogen®). Visualization was possible using UV transilumi- The thermocycler was programmed with 29 cycles, each one with
nator (Vilbert Loumart®) and was photo documented (Sony® Graphic denaturation at 95 °C for 2 min, 57 °C as annealing temperature and 72 °C
Printer UP-895CE). for 2.5 min. The samples were maintained cooled at 4 °C until their
withdrawal. The program specifications for Typhimurium serotype were
2.3. Positive controls for PCR basically the same, with the exception of the annealing temperature,
which was established at 55 °C. All PCR reactions were visualized at 2.0%
Bacteria genus Salmonella and its serotypes Enteritidis, Typhi and (w/v) agarose gel in UV transiluminator and photo documented.
Typhimurium were only identifiable after the selection of the specific Specificity tests were performed to exclude the possibility of
regions of each serotype's genes. Serotype Agona was used for genus interaction between the pairs of primers. It was also verified that
identification. A fragment of 204 base pairs (bp) of the OMPC gene nonspecific reactions could be possible between those and the DNAs
was selected. This gene is responsible for protein C involved in the of each serotype when reactions were performed with primers not
invasion of epithelial cells (OMPCF and OMPCR primers). A 304 bp specific for the DNA as a template.
fragment of the SdfI gene, chromosomal region related to invasiveness The PCR products were submitted to automatic sequencing to
and infection of poultry and eggs (ENTF and ENTR primers) was ensure the fidelity of transcription. MegaBACE®1000 (Amersham
selected for serotype Enteritidis. For Typhimurium serotype, a 401 bp Biosciences®) was the equipment used. For the sequencing procedure
Spy gene fragment was selected, which encodes a specific periplas- two 0.5 mL polypropylene tubes with PCR products of Salmonella sp.
matic protein (TyphF and TyphR primers). All theses primers were and Enteritidis, Typhi and Typhimurium serotypes were prepared.
compiled from Alvarez et al. (2004). Concerning serotype Typhi, a Each of the polypropylene tubes contained the respective gene region
fragment of 738 bp ViaB gene, coding for bacterial capsule protein amplified for each serotype. Those products were purified with the
synthesis (Vi antigen) was selected (ViaBF and ViaBR primers). This PCR purification KIT product (QIAquick 50, Quiagen®). All reactions
pair of primers was described in Kumar et al. (2006). The present were made with forward primer only and another reaction with
sequences of primers are shown in Table 1. The primers were obtained reverse primer for each PCR serotype reaction.
from IDT® — Integrated DNA Technologies (Prodimol®). The obtained sequences were compared with sequences stored in
For each pair of designed primers, PCR reactions were performed GenBank database using BLAST — Basic Local Alignment Search Tool
with total DNA extracted from strains of Salmonella sp. and serotypes (available at www.ncbi.nlm.nhm).
Enteritidis, Typhi and Typhimurium (FIOCRUZ positive controls) for
confirmation of the primers' specificity. Protocols for each PCR 2.4. Positive controls for multiplex PCR
reaction respected the particularities of each pair of designed primers,
regarding their annealing temperatures. The PCR reactions were After the establishment of the individual protocols for the genus
performed with different concentrations of Magnesium Chloride Salmonella and the selected serotypes, various multiplex PCR (mPCR)
 C.G. de Freitas et al. / International Journal of Food Microbiology 139 (2010) 15–22
were performed with total DNA of positive controls. These reactions
consisted of adding all pairs of primers (Salmonella sp, serotypes
Enteritidis, Typhi and Typhimurium) in a single PCR reaction, added
with the reagent mix previously prepared. As described, the reaction
has the pair of primers that amplifies InvA gene (OMPF, OMPCR —
204 bp), and specific primers for each of the serotypes Enteritidis
(ENTF, ENTR — 304 bp), Typhi (ViaBF, ViaBR — 738 bp) and
Typhimurium (TyphF, TyphR — 401 bp). It was expected to observe
two fragments in 2.0% agarose gel: one corresponding to the genus
(204 bp) and another to identify the serotypes Enteritidis (304 bp),
Typhi (738 bp) or Typhimurium (401 bp).
2.5. Contamination of poultry meat broth
In order to prove the viability of the proposed Salmonella sp.
identification technique, poultry meat broth was artificially inocu-
lated with the same strains of each serotype to obtain different
gradients of bacteria concentration. The objective was to verify if
there was any inhibitory factor in this broth resulting from the
processing of the samples, which could interfere within the
expected result. This step also aimed at checking the sensitivity of
PCR related to bacteria concentration in the medium. Poultry meat
broth contamination was performed as described by Kumar et al.
(2006). A poultry meat broth was made using 25 g of cooled chicken
meat that previously tested negative for Salmonella sp., added with
250 mL of buffered Peptone Water Broth (Merck®) and homoge-
nized in a sterile stomacher (Mayo®) bag for 1 min. Five aliquots of
50 mL were made in Falcon® tubes, which were then subjected to
centrifugation (Eppendorf Centrifuge 5804®) for 5 min at 3000 rpm;
then the supernatants were removed and stored individually under
refrigeration. Five aliquots of 100 mL of 0.85% sterile saline solution
were inoculated with Salmonella serotypes Agona, Enteritidis, Typhi
and Typhimurium bacteria to achieve the concentration of 10− 1,
100, 101, 102 and 103 CFU/mL. Colonies of Salmonella sp. were
counted and their numbers were reported as colony forming units
(CFU). Then these solutions were filtered on Milipore® filter
(0.45 µm of diameter). The filters were inoculated separately in
50 mL of supernatant of the poultry meat broth at 24 °C for 24 h.
After this period, total DNA extraction was carried out with 1 mL of
the broth to be used for PCR and mPCR reactions.
2.6. Origin of chilled poultry carcasses and viscera
The poultry carcasses and viscera were collected from commercial
establishments in the Federal District, Brasília, Brazil. A total of 200
samples were analyzed, of which 127 were poultry carcasses and 73 of
its viscera (composed by heart, gizzard and liver with about 100 g of
total final weight). Those viscera were collected due to their common
habit of consumption in Brazil.
Samples were transported in isothermal boxes to the Laboratory of
Food Microbiology (LAMAL), University of Brasília, where they were
immediately weighted to form a 25 g sample in sterile laminar flow
and packed in stomacher bags. These were added to 225 mL of
Buffered Peptonated Water, homogenized and incubated at 37 °C for
24 h. The remaining sample material was kept under freezing.
2.7. Total DNA extraction of poultry carcasses and viscera
The search for Salmonella sp. was based on a total of 25 g composed
of various parts of each sample, which were inoculated in 225 mL of
enrichment culture buffered peptone water broth (BPW; Merck®).
which allows Salmonella's growth, according to Normative Instruction
n°62, September 18, 2003, of the Ministry of Agriculture, Livestock
and food Supply (BRASIL, 2003a,b).The material was incubated in a
bacteriological incubator (Quimis®) at 37 °C for 24 h, and then 1 mL
was collected for the extraction of total DNA, which, in turn, followed
17
the phenol, chloroform, isoamyl alcohol method described by
Sambrook and Russell (2006).
2.8. PCR of poultry carcasses and viscera
The total DNA obtained from each sample was subjected to PCR
with OMPCF, OMPCR primers for genus Salmonella identification.
Samples that tested positive (204 bp fragment in 2.0% w/v agarose
gel), were subsequently tested with multiplex PCR for serotype
identification — Enteritidis, Typhi and Typhimurium.
2.9. Multiplex PCR of poultry carcasses and viscera
The mPCR was performed in positive samples for Salmonella sp.
according to the previously described protocol for the positive
controls. In the same mPCR were added OMPCF, OMPCR primers
(Salmonella sp.), ENTF, ENTR primers (S. Enteritidis) and ViaBF and
ViaBR primers (S. Typhi), all of them with 57 °C annealing temper-
ature. There was also a second mPCR with OMPCF, OMPCR primers
and TyphiF and TyphR primers (S. Typhimurium) due to the annealing
temperature of the last pair of primers being 55 °C, which does not
interfere in the genus Salmonella identification by OMPCF, OMPCR
primers.
2.10. Microbiological analyses of poultry carcasses and viscera
In parallel with PCR and mPCR identification for the presence of
Salmonella sp. and its serotypes, microbiological research was also
performed as indicated by the Normative Instruction N°62 of Ministry
of Agriculture, Livestock and Food Supply (BRASIL, 2003a,b) to
confirm whether the obtained results were compatible with those
from mPCR. These analyses were performed by pre-enrichment broth
(BPW, 24 h at 37 °C), selective enrichment (Rappaport Vassiliadis —
24 h at 42 °C; Merck®). The samples were immediately subcultured
onto Salmonella–Shigella agar (Merck ® ) and McConkey agar
(Merck®). Colonies were identified on the basis of biochemical tests
and confirmed as Salmonella sp. by Probac® serological test — Somatic
(O) and flagellar (H) antigens.
2.11. Estimation of the frequency of Salmonella sp. in poultry carcasses
and viscera, in the Federal District, Brazil
Each sample of poultry carcasses (n = 127) and viscera (n = 73)
were classified as positive for Salmonella sp. when both mPCR and
traditional bacteriological methods were positive. The frequency
distribution of test-positive samples was modeled using the binomial
process. Thus, there was no assumption of data being normally
distributed. The uncertainty about the frequency (p) true value was
quantified using a beta distribution, with a uniform prior for p, due to
the lack of prior information on this parameter (Vose, 2008).
Calculations were carried out using the software @Risk 4.5 for
Microsoft Excel (Palisade Corporation®).
3. Results
3.1. Extraction of total DNA of positive controls
The genetic material extracted from isolates (serotypes Typhi,
Enteritidis and Typhimurium) were quantified using high mass
marker (Invitrogen®) in 0.8% (w/v) agarose gel. The observed
quantities of total DNA ranged between 5–10 ng/µL.
3.2. PCR of positive controls
The Polimerase Chain Reaction (PCR) was performed for genus
Salmonella and for each serotype Enteritidis, Typhi and Typhimurium.
18 C.G. de Freitas et al. / International Journal of Food Microbiology 139 (2010) 15–22
The aim was to standardize them individually. The optimum MgCl2
concentration in which the reaction occurred in a more adjusted way
was 3.0 mM. Concentrations remained in the initial proposals for the
other reagents: 10 pmol of each forward or reverse primer; 2.0 mM of
dinucleotide (dNTP), the final concentration of 1× of 10× buffer; 2.0 U
of Taq polymerase, 5.0 ng of DNA template and sterile MilliQ water to
complete 25 µL of reaction.
Each of the used primers had different annealing temperature
(Tm). In order to perform mPCR, the considered annealing temper-
ature was an average for all pairs of primers 57 °C for identification of
Salmonella sp. and serotypes Enteritidis and Typhi, and 55 °C for
Typhimurium serotype. These temperatures were considered appro-
priate to mPCR reactions. In each performed reaction, were observed
the DNA amplification identifying Salmonella sp. (304 bp) and its
respective serotype in each case: 304 bp for serotype Enteritidis,
738 bp for serotype Typhi and 401 bp for serovar Typhimurium
(Fig. 1).
3.3. Results of automatic sequencing
After PCR reactions, purification of the products was performed
and sent to MegaBACE® automatic sequencing. The results demon-
strated the transcription fidelity and specificity (range around 98% of
transcription fidelity) of used primers and were also compared with
the GenBank database by BLAST program. It demonstrated that the
obtained sequences were reliable both for identification of Salmonella
sp. and for each serotypes Enteritidis, Typhi and Typhimurium.
3.4. Specificity of the primers in the positive controls
Specificity tests were made between all pairs of primers. There
were several individual PCR reactions by adding DNA from one of the
serotypes of Salmonella sp. as template and pairs of primers of other
serotypes. The results of these reactions demonstrated that the
primers were specific to target the regions of their respective
serotypes. No nonspecific reactions occurred with other serotypes
primers.
3.5. Multiplex PCR of positive controls
The stardardization of multiplex PCR primarily involved the
optimization of PCR for Salmonella sp. and serotypes Enteritidis,
Typhi and Typhimurium separately. Thus, it was observed the
minimal optimal concentrations of each of the reagents for PCR
reaction.
Fig. 1. PCR of positive controls for genus Salmonella and its serotypes Enteritidis, Typhi
and Typhimurium. 1.1) 100 bp DNA ladder (Promega®); 1.2) Fragment of 204 bp of
Salmonella sp.; 1.3) 304 bp fragment of serotype Enteritidis; 1.4) 738 bp fragment of
serotype Typhi and 1.5) 401 bp fragment from serotype Typhimurium. 2.0% (w/v)
agarose gel added with 0.02% (w/v) of ethidium bromide.
The mPCR reaction was prepared with all the pairs of primers at a
concentration of 10 pmol (OMPCF, OMPCR for Salmonella sp.; ENTF,
ENTR for serotype Enteritidis; ViaBF, ViaBR for serotype Typhi and
TyphiF, TyphR for serotype Typhimurium). The mPCR was made with
the same concentrations of PCR reactions. The final reaction volume,
however, was 30 µL. The mPCR was successfully performed, not
experiencing any nonspecific reaction (Fig. 2).
3.6. Contaminations of poultry meat broth
The contaminations of poultry meat broth with different bacteria
concentrations (10− 1, 100, 101, 102 and 103 CFU/mL) were success-
fully performed as all PCR reactions worked with different serotypes
(Fig. 3-A, 3-B, 3-C and 3-D). No inhibitory substances in poultry meat
broth was observed. The mPCR technique worked in all proposed
concentrations of bacteria with sensitivity of detection in quantities of
up to 10− 1 CFU/mL.
3.7. Extraction of total DNA of poultry carcasses and viscera sold in the
Federal District
The total DNA from each sample was successfully obtained from
the pre-enrichment broth at concentrations that ranged approxi-
mately between 5–10 ng/µL. The total DNA of each bacteria was used
as templates in PCR and mPCR reactions for Salmonella sp. and
serotypes Enteritidis, Typhi and Typhimurium research.
3.8. Detection of Salmonella sp. and serotypes Enteritidis, Typhi and
Typhimurium by mPCR and by traditional microbiological methods in
chilled poultry carcasses and viscera sold in the Federal District
Out of 73 samples of viscera, two samples tested positive for
Salmonella sp. in PCR reaction. The same samples were then submitted
to the mPCR reaction for serotype identification. One of the samples
was identified as S. Enteritidis (Fig. 4-A) whereas the serotype of the
other sample could not be identified by mPCR (Fig. 4-B). This is
acceptable because this project worked with three serotypes and
Salmonella sp. comprises about 2557 known serotypes. None of the
127 poultry carcasses were positive for Salmonella sp, even though
fragments of the neck skin, below the wings and near cloacal region of
each sample were collected, as recommended by Nógrády et al.
(2008).
The same two samples of chilled viscera that tested positive for
mPCR yielded positive results in conventional microbiological
Fig. 2. Multiplex PCR for identification of Salmonella sp. and serotypes Enteritidis, Typhi
and Typhimurium. 2.1) 100 bp DNA ladder (Promega®); 2.2) fragment of 204 bp of the
genus Salmonella; 2.3) fragment of 204 bp of the genus Salmonella and 304 bp fragment
of the serotype Enteritidis; 2.4) fragment of 204 bp of the genus Salmonella and 738 bp
fragment of the serovar Typhi; 2.5) fragment of 204 bp of the genus Salmonella and
401 bp fragment of the serovar Typhimurium. 2.0% (w/v) agarose gel with 0.02% (w/v)
of ethidium bromide.
 C.G. de Freitas et al. / International Journal of Food Microbiology 139 (2010) 15–22 19

Fig. 3. 3-A: Result of mPCR for Salmonella sp. detection in poultry meat broth previously contaminated with Salmonella enterica serotype Agona in concentrations of 10− 1, 100, 101,
102 and 103 CFU/mL. A.1) 100 bp DNA ladder (Promega®); A.2) negative control; A.3) contaminated with 10− 1 CFU/mL; A.4) contaminated with 100 CFU/mL; A.5) contaminated
with 101 CFU/mL; A.6) contaminated with 102 CFU/mL; A.7) contaminated with 103 CFU/mL. 2.0% (w/v) agarose gel with 0.02% (w/v) of ethidium bromide. 3-B: Result of mPCR for
the genus Salmonella sp. detection in poultry meat broth previously contaminated with Enteritidis in concentrations of 10− 1, 100, 101, 102 and 103 CFU/mL. B.1) 100 bp DNA ladder
(Promega®); B.2) negative control; B.3) contaminated with 10− 1 CFU/mL; B.4) contaminated with 100 CFU/mL; B.5) contaminated with 101 CFU/mL; B.6) contaminated with
102 CFU/mL; B.7) contaminated with 103 CFU/mL. 2.0% (w/v) agarose gel with 0.02% (w/v) of ethidium bromide. 3-C: Result of mPCR for the genus Salmonella sp. detection in poultry
meat broth previously contaminated with Enteritidis and Typhi in concentrations of 10− 1, 100, 101, 102 and 103 CFU/mL. C.1) 100 bp DNA ladder (Promega®); C.2) negative control;
C.3) contaminated with contamination 10− 1 CFU/mL; C.4) contaminated with 100 CFU/mL; C.5) contaminated with 101 CFU/mL; C.6) contaminated with 102 CFU/mL; C.7)
contaminated with 103 CFU/mL. 2.0% (w/v) agarose gel with 0.02% (w/v) of ethidium bromide. 3-D: Result of mPCR for the genus Salmonella sp. detection in poultry meat broth
previously contaminated with serotypes Enteritidis and Typhimurium in concentrations of 10− 1, 100, 101, 102 and 103 CFU/mL. D.1) 100 bp DNA ladder (Promega®); D.2) negative
control; D.3) with contamination 10− 1 CFU/mL; D.4) contaminated with 100 CFU/mL; D.5) contaminated with 101 CFU/mL; D.6) contaminated with 102 CFU/mL; D.7) contaminated
with 103 CFU/mL. 2.0% (w/v) agarose gel with 0.02% (w/v) of ethidium bromide.

methods, using biochemical tests. The latter were not able to identify
the serotype. There was also perfect agreement between mPCR and
microbiological methods in negative samples. None of the 127 poultry
carcasses were positive to Salmonella sp. in microbiological tests,
therefore, in agreement with mPCR results. Although, this result was
expected because there was a small number of poultry carcasses.
3.9. Frequency distribution of Salmonella sp. in poultry viscera, in the
Federal District
Two samples of poultry viscera tested positive both for mPCR and
microbiological tests, out of 73 samples. Fig. 5 displays the frequency
distribution that results from these parameters applied to a beta

Fig. 4. (A) Multiplex PCR for identification of Salmonella sp. and serotype Enteritidis from cooled poultry viscera. A.1) 100 bp DNA ladder; A.2) mPCR fragment of 204 bp of the genus
Salmonella and 304 bp fragment of the serotype Enteritidis. 2.0% (w/v) agarose gel with 0.02% (w/v) of ethidium bromide. (B) Multiplex PCR for identification of Salmonella sp. from
chilled poultry viscera. B.1) 100 bp DNA ladder; B.2) mPCR fragment of 204 bp of the genus Salmonella sp. 2.0% (w/v) agarose gel with 0.02% (w/v) of ethidium bromide.
20 C.G. de Freitas et al. / International Journal of Food Microbiology 139 (2010) 15–22
Fig. 5. Frequency distribution of Salmonella spp. contamination in chilled poultry
viscera in the Federal District, Brazil, based on a beta distribution.
distribution, in order to describe the uncertainty about the true value
of test-positive samples. The 95% confidence interval ranged from
0.84% to 9.42%, with the mean at 4.0%, the mode at 2.74% and median
3.6%.
4. Discussion
In the present work we chose phenol, chloroform and isoamyl
alcohol for extracting total DNA of samples, in order to minimize
possible nonspecific reactions or even the absence of reaction, as
discussed by Sambrook and Russell (2006). This procedure was
different from those described elsewhere. Kumar et al. (2006) used
the boiling method for total DNA extraction. Alvarez et al. (2004),
Halatsi et al. (2006) had also opted for this method, although these
authors reported no problems due to their extraction method. The
optimum annealing temperature of 57 °C proved to be adequate for
genus Salmonella and Enteritidis and Typhi serotype identification,
preventing nonspecific reactions. Our results of annealing tempera-
ture were similar to those found by Alvarez et al. (2004); these
authors also used the temperature of 57 °C for Salmonella sp. and for
serotypes Enteritidis and Typhimurium. However, in our work the
temperature of 57 °C didn't occur for S. Typhi, that was of 55 °C. As to
the latter, the results of this study are consistent with those obtained
by Kumar et al. (2006), who reported an annealing temperature of
55 °C.
With regard to the specificities of PCR reactions, a specificity of 98%
of transcription for each serotype was observed, which is consistent
with the reactions obtained by Alvarez et al. (2004), who worked with
Salmonella sp. and Enteritidis and Typhimurium serotypes. Our results
were also similar to those reported by Kumar et al. (2006), who
worked with S. Typhi. Salmonella sp. detection in poultry meat broth,
as suggested by Kumar et al., 2006, with strains of serotype Typhi from
Madhya Pradesh in India, that was achieved successfully with
Salmonella sp. and serotypes Enteritis, Typhi and Typhimurium strains
from Brazil in all different gradients of bacterial concentration. Our
results were similar to the results showed by Lee et al. (2009). They
detected the serotypes Enteritidis and Typhimurium in pork and beef
by multiplex real-time PCR.
Salmonella sp. detection both by mPCR and conventional micro-
biological methods yielded similar results in poultry carcasses and
viscera. Two samples of chilled viscera were confirmed positive for
genus Salmonella, similar to what has been reported by the
Monitoring Program of the Prevalence of Bacterial Resistance in
Chickens of the National Health Surveillance Agency — ANVISA
(Prebaf, 2008), which found 2.58% positive Salmonella sp. strains in
the Federal District. These data are also similar to those of the National
Prevalence of Salmonella sp. in chicken meat, which was 4.0%, exactly
the mean of the distribution displayed in Fig. 5. Serotype identification
was made by mPCR, which identified S. Enteritidis (1.37%) in poultry
viscera in the Federal district. This result was similar to the findings of
Fuzihara et al. (2000). They reported that the serotype Enteritidis was
the most prevalent in carcasses of poultry meat, using microbiological
methods. These data are not in agreement with the Monitoring
Program of the Prevalence of Bacterial Resistance in Chickens of the
National Health Surveillance Agency (ANVISA) that is (48.8%) because
this is a National result, and this program doesn't refer about the
serotype identification in the region of the Federal District. Even on
the national identification of serotype Enteritidis, we emphasized that
the methodology used for this differentiation was performed by
FIOCRUZ, which is a reference laboratory for such analysis in Brazil.
The advantage of the adapted technique developed in the present
work is that it identified not only Salmonella sp. but also Enteritidis
serotype by mPCR, which reduces the diagnostic time and also
rendered it possible to single out the three most common serotypes in
the region.
Nunes et al. (1995) used traditional microbiological methods to
conduct a survey of Salmonella sp. in poultry carcasses and its cuts, in
chickens sold in Goiânia, Brazil, in 1995. These researchers used
carcass wash and skin maceration and found an average frequency of
13.2% of Salmonella sp. They observed Brandenburg as the most
common serotype. These results might have been explained by the
fact that Normative Instruction n 70 of the Ministry of Agriculture,
Livestock and Food Supply (Brasil, 2003b) was not yet in place at that
time.
Multiplex PCR results gave two positive results for Salmonella sp.
in poultry viscera, with one of the samples identified as S. Enteritidis
by mPCR. This technique was successfully used by Kumar et al.
(2006) for the detection of different virulence genes in serotype
Typhi. Alvarez et al. (2004) used the same technique for the
detection and epidemiological typing of Salmonella sp. in clinical
samples of patients.
The identification of multiple serotypes of Salmonella sp. by
mPCR as proposed by this work is similar to previous work done by
O'Regan et al. (2008), who used real-time mPCR for multiple
Salmonella sp. serotype identification in poultry meat. However, the
technique they used was more expensive than ours. In addition to
the serotypes Enteritidis and Typhimurium proposed in this work,
those researchers also worked with Gallinarum, Kentucky and
Dublin serotypes. The present study worked also with the Typhi
serotype due to its relevance to public health. Nógrády et al. (2008),
in Hungary, diagnosed the Infantis serotype in slaughterhouse
samples and in clinical cases of human salmonellosis. These authors
identified the serotype Infantis in 43% of environmental samples
and 35% of contaminated poultry meat (carcasses and chicken cuts).
The same study found Salmonella sp. by PCR in 5 samples, which
were negative to microbiological tests, which seems to indicate that
Salmonella sp. may be present in quantities below the detection
limit of microbiological research. Since PCR has DNA as the
diagnostic target, it doesn't depend on the presence of significant
amounts of bacteria in food.
The use of molecular tools such as Polymerase Chain Reaction
(PCR) and mPCR improves the quality of the diagnostic procedures
and mPCR enhances microbiological quality control. Since mPCR is
able to detect Salmonella sp. within hours, industries and producers
are able to take corrective measures in order to avoid contamination
of their products. The successful employment of mPCR technique, as
proposed by the authors, for Salmonella sp. and Enteritidis, Typhi and
Typhimurium serotypes, is an encouraging development, which may
 C.G. de Freitas et al. / International Journal of Food Microbiology 139 (2010) 15–22
be applied to other serotypes. The technique can be applied for
Salmonella sp. identification in other meat bases.
The observed presence of Salmonella sp. in poultry viscera but not
in carcasses, is consistent with the findings of Nunes et al. (1995), who
found a higher frequency of enterobacteria in chicken cuts compared
to carcasses. Therefore, we can infer that manipulation of poultry
viscera and chicken cuts may be associated with a greater likelihood of
contamination with Salmonella sp.
The presence of Salmonella sp. and Enteritis and Typhimurium
serotypes in foods of animal origin, is indicative of flaws in poultry
handling or mishandling of food, as these two serotypes may affect
both humans and chicken. The presence of S. Typhi suggests food
contamination by an asymptomatic carrier of this serotype. Badrinath
et al. (2004) investigating a salmonellosis outbreak by S. Enteritidis in
a Chinese restaurant in Sulfolk, England, used the data collected by the
epidemiological surveillance to establish the likely source of contam-
ination. Also related to the proper handling of food and its importance
to public health, Beatty et al. (2009) reported an outbreak of
salmonellosis in a conference held in Dallas, Texas in 2002. The risk
analysis showed the handling of food by asymptomatic carriers of
Salmonella sp. serotype Enteritidis as a source of infection.
The mPCR technique for Salmonella sp. and Enteritidis, Typhi and
Typhimurium serotypes can be used in the evaluation of naturally
contaminated foods as shown by our results. The technique also
solves some limitations of the conventional microbiological isola-
tion, namely the time to identify the organism, which ranges from
1 week to 15 days, in contrast with 18 h for PCR or mPCR.
Furthermore, the fact that there is no need of a significant number
of viable cells for detection of bacteria is another advantage, since
the PCR can generate millions of copies from a strip of DNA.
However, it is important to emphasize that PCR cannot distinguish
between dead and living cells because this technique uses just DNA
as template. So, it must make the PCR test in parallel with the
microbiological test. Furthermore it must be considered that
differences between PCR and microbiological test results may be
attributable to the fact that target cells may have been injured or
died despite the enrichment procedures that were non-culturable
but detectable by PCR.
5. Conclusions
The PCR and mPCR technique were successfully adapted to identify
Salmonella sp. and Enteritis, Typhi and Typhimurium positive controls.
The protocols and primers used were effective for the amplification of
a fragment of 204 bp of ompC gene for the genus Salmonella, a
fragment of 304 bp of the SdfI gene of serotype Enteritidis, a fragment
of 738 bp of the ViaB gene, a fragment of serotype Typhi and 401 bp of
the gene Spy of serotype Typhimurium.
The results yielded by molecular detection and conventional
microbiology were in perfect agreement, both for positive and
negative Salmonella sp samples. The mPCR identified, however, the
serotype Enteritidis in a sample of cooled poultry viscera, which could
not be achieved by biochemical tests. In the other positive sample of
chilled poultry viscera, mPCR identified only the genus Salmonella. It is
worth emphasizing that the project worked with three of the known
2557 serotypes of Salmonella enterica.
Given the importance of foodborne outbreaks in public health, this
work presents an alternative diagnostic procedure for the identifica-
tion of pathogens. Its reliability and execution time makes it possible
to be used in conjunction with conventional microbiological methods
or in the event of an outbreak. The technique described here should be
further studied and refined and protocols for the identification of
other serotypes ought to be developed. The basis of the methodology
can be extended to studies of emerging serotypes in public health, as
well as to other microorganisms of importance in food microbiology.
21
References
Aarestrup, F.M., et al., 2007. International spread of multidrug-resistant Salmonella
Schwarzengrund in food products. Emerging Infectious Diseases 13 (5), 726–731.
ABEF. Produção mundial de carne de frango. Site of Access: www.abef.com.br. Acessed
in 05th Oct 2008.
Agunos, A., et al., 2007. Effect of dietary beta1-4 mannobiose in the prevention of
Salmonella enteritidis infection in broilers. British Poultry Science 48 (3), 331–341.
Alizadeh, A.H., et al., 2007. Escherichia coli, Shigella and Salmonella species in acute
diarrhoea in Hamedan, Islamic Republic of Iran. Eastern Mediterranean Health
Journal 13 (2), 243–249.
Alvarez, J., et al., 2004. Development of a multiplex PCR technique for detection and
epidemiological typing of Salmonella in human clinical samples. Journal of Clinical
Microbiology 42 (4), 1734–1738.
Arya, S.C., Agarwal, N., 2008. Typhoid fever vaccines: systematic review and meta-
analysis of randomised clinical trials. Vaccines 26 (3), 291.
Badrinath, P., et al., 2004. An outbreak of Salmonella Enteritidis phage type 34a infection
associated with a Chinese restaurant in Suffolk United Kingdom. BMC Public Health
4, 40.
Beatty, M.E., et al., 2009. Large Salmonella Enteritidis outbreak with prolonged
transmission attributed to an infected food handler. Epidemiology and Infection
137 (3), 417–427.
Bjerrum, L., et al., 2005. The influence of whole wheat feeding on Salmonella infection
and gut flora composition in broilers. Avian Disease 49 (1), 9–15.
Blixt, O., et al., 2007. Pathogen specific carbohydrate antigen microarrays: a chip for
detection of Salmonella O-antigen specific antibodies. Glycoconjugate Journal 25
(1), 27–36.
BRASIL. Normative Instruction number 62 de 18 de setembro de 2003. Métodos
Analíticos Oficiais para Análises Microbiológicas para Controle de Produtos de
Origem Animal e Água. Ministério da Agricultura, Pecuária e Abastecimento 2003
Site of Access: www.agricultura.gov.br.
BRASIL. Normative Instruction number 70 de 10th october, 2003. Programa de Redução
de Patógenos – Monitoramento Microbiológico e Controle Sanitário de Salmonella
sp. em Carcaças de Frangos e Perus. Departamento de Defesa e Inspeção Vegetal.
Secretaria de Defesa Agropecuária. Ministério da Agricultura, Pecuária e Abaste-
cimento 2003b Site of access: www.agricultura.gov.br.
Chaicumpa, W., et al., 1995. Monoclonal antibody-based dot-blot ELISA for the
detection of Salmonella in foods. Asian Pacific Journal of Allergy and Immunology
13 (2), 159–166.
Che, Y.H., et al., 2000. Rapid detection of Salmonella typhimurium in chicken carcass
wash water using an immunoelectrochemical method. Journal of Food Protection
63 (8), 1043–1048.
Chen, S., et al., 1997. The evaluation of a fluorogenic polymerase chain reaction assay for
the detection of Salmonella species in food commodities. International Journal of
Food Microbiology 35 (3), 239–250.
Christensen, J.P., et al., 1997. Salmonella enterica serovar Tennessee infections in
broilers. Avian Pathology 26 (1), 155–168.
Deseo, L., Engeli, P., 1979. Symptomless enteritis—Salmonella excreters in an abbatoir.
Schweizerische Medizinische Wochenschrift. Journal Suisse de Medecine 109 (50),
1995–1999.
Fach, P., et al., 1999. Evaluation of a polymerase chain reaction-based test for detecting
Salmonella spp. in food samples: Probelia Salmonella spp. Journal of Food Protection
62 (12), 1387–1393.
Foley, S.L., et al., 2007. Comparison of molecular typing methods for the differentiation
of Salmonella foodborne pathogens. Foodborne Pathogens and Disease 4 (3),
253–276.
Fuzihara, T.O., et al., 2000. Prevalence and dissemination of Salmonella serotypes along
the slaughtering process in Brazilian small poultry slaughterhouses. Journal of Food
Protection 63 (12), 1749–1753.
Gallegos-Robles, M.A., et al., 2009. PCR detection and microbiological isolation of
Salmonella spp. from fresh beef and cantaloupes. Journal of Food Science 74 (1),
M37–M40.
Halatsi, K., et al., 2006. PCR detection of Salmonella spp. using primers targeting the
quorum sensing gene sdiA. FEMS Microbiology 259, 201–207.
Kaku, M., et al., 2000. Alimentar por Salmonella Enteritidis no Noroeste do Estado de São
Paulo, Brasil. Revista de Saúde Pública 29 (2), 127–131.
Kornschober, C., et al., 2009. Salmonellosis in Austria: situation and trends. Wiener
Klinische Wochenschrift 121 (3–4), 96–102.
Kottwitz, L.B.M., et al., 2008. Contaminação por Salmonella spp. em uma cadeia de
produção de ovos de uma integração de postura commercial. Arquivos Brasileiro de
Medicina Veterinária e Zootecnia 60 (2), 496–498.
Kumar, S., et al., 2006. Detection of Salmonella enterica serovar Typhi (S. Typhi) by
selective amplification of invA, viaB, fliC-d and prt genes by polymerase chain
reaction in mutiplex format. Letters in Applied Microbiology 42 (2), 149–154.
Kumar, R., et al., 2008. Molecular fingerprinting of Salmonella enterica subsp. enterica
Typhimurium and Salmonella enterica subsp. Enterica Derby isolated from tropical
seafood in South India. Molecular Biotechnology 40 (1), 95–100.
Lee, S.H., et al., 2009. A multiplex real-time PCR for differential detection and
quantification of Salmonella spp., Salmonella enterica serovar Typhimurium and
Enteritidis in meats. Journal Veterinary Science 10 (1), 43–51.
Malkawi, H.I., Gharaibeh, R., 2003. Multiplex PCR for the direct detection of Salmonella
enterica from chicken, lamb and beef food products. Journal of Basic Microbiology
43 (4), 328–336.
Muth, M.K., et al., 2009. Analysis of Salmonella control performance in U.S. young
chicken slaughter and pork slaughter establishments. Journal of Food Protection 72
(1), 6–13.
22 C.G. de Freitas et al. / International Journal of Food Microbiology 139 (2010) 15–22
Nógrády, N., et al., 2008. Prevalence and characterization of Salmonella Infantis isolates
originating from different points of the broiler chicken–human food chain in
Hungary. International Journal of Food Microbiology 127, 162–167.
Nunes, I.A., et al., 1995. Ocorrência de Salmonella sp. em carcaças e cortes de frangos
comercializados em Goiânia-GO. Anais da Escola de Agronomia e Medicina
Veterinária da Universidade Federal de Goiás 25 (2), 1–5.
Ochiai, R.L., et al., 2008. A study of typhoid fever in five Asian countries: disease burden
and implications for controls. Bulletin of the World Health Organization 86 (4),
260–268.
O'Regan, E., et al., 2008. Development of a real-time multiplex PCR assay for the
detection of multiple Salmonella serotypes in chicken samples. BMC 21 (8), 156.
Prebaf. Programa Nacional de Monitoramento da Prevalência e da Resistência
Bacteriana em Frango – Relatório do monitoramento e da prevalência e do perfil
de suscetibilidade aos antimicrobianos em Enterococcus e Salmonella isolados de
carcaças de frango congeladas comercializadas no Brasil. Agência Nacional de
Vigilância Sanitária – ANVISA, Ministério da Saúde, 2008; 186p. Site of Access:
www.anvisa.gov.br.
Rabsch, W., et al., 2001. Non-typhoidal salmonellosis: emerging problems. Microbes
and Infection 3 (3), 237–247.
Rasschaert, G., et al., 2008. Contamination of carcasses with Salmonella during poultry
slaughter. Journal of Food Protection 71 (1), 146–152.
Sambrook, J., Russell, D.W., 2006. The Condensed Protocols from Molecular Cloning: a
Laboratory Manual. Cold Spring Harbor Laboratory Press.
Vose, D., 2008. Risk AnalysisThird Edition. John Wiley and Sons, England.